每日：科技领域之外的AI机遇

Podcast: Signal over Noise | AI trade at a crossroads, on Apple or Spotify (6 mins)Video: Top of Mind in APAC | The current bull market and risks to watch (5 mins)Video: Market Playbook | Why investors should revisit Europe (4 mins)

A reassessment of the AI trade and the unwinding of leveraged positions under pressure have led to a peak-to-trough decline of 28.6% in the Philadelphia Semiconductor Index over the past six weeks. While sentiment has improved slightly in recent days, we believe the next phase of the AI growth story requires more than just owning companies linked to AI spending.

We have highlighted our preference for a selective, balanced approach to tech exposure, but investors can also consider AI opportunities from a thematic angle. In fact, since the technology is transforming businesses and industries, we see opportunities across sectors that are helping enable, support, or apply AI at scale.

Power constraints underscore the investment required to meet rising demand. Access to power is one of the biggest constraints in AI development, and it is also one of the areas with the least substitution risk, in our view. Power generation capacity and grid connections face long lead times, and chips, which have shorter life cycles, can be affected by the long cycles required to make the equipment that produces the chips. This underpins our positive view on the Power and resources theme, which focuses on opportunities related to power infrastructure and industrial technology, both essential parts of the value chain for developing AI and satisfying rising electricity demand.

AI demand continues to support semis. With hyperscalers raising guidance for their near-term capital spending commitments, we continue to see beneficiaries in the semiconductor and hardware segment of the AI value chain. But we think selectivity is key, and we see the biggest opportunity now in semicap equipment. We expect supply bottlenecks to be the most prominent for semicap equipment, and forecast revenue will increase at an annual compound growth rate of roughly 30% over the next three years. We also believe foundries offer compelling risk-reward, as oversupply risks remain manageable, valuations are reasonable, and pricing power is improving. Compute, meanwhile, should continue to benefit from the rising adoption of agentic AI.

Health care is a promising AI application. We have long held the view that AI applications are the battleground where most value will be created over the long term, and among those applications, health spending is one of the most price-inelastic goods. We therefore believe pharma and biotech companies with proprietary data assets offer a unique opportunity. In fact, AI-led advances in the health care sector have been notable: Lila’s latest mRNA therapeutics delivered performance twice as effective as conventional technologies, while Insilico Medicine’s AI-discovered drug to treat idiopathic pulmonary fibrosis has entered Phase III clinical trials. We continue to favor health care globally, and believe the sector offers opportunities backed by the structural trend of Longevity.

So, in addition to holding tech stocks and without taking any single-security views, we believe investors can also position for AI growth by broadening their exposure to industrials, utilities, and health care.